This study is a randomized controlled trial of a multilevel social-behavioral intervention for HIV-positive injecting drug users (IDUs) in Vietnam. Our ongoing NIMH-funded study at the same site found high HIV prevalence (42%) among IDUs. Among HIV-positive IDUs, 64% percent shared injecting equipment and 65% of sexually active IDUs engaged in unprotected sex in the past year. The intervention draws on social action, social identity and diffusion of innovation theories. At the structural level, the intervention consists of 2 community-based stigma reduction programs;at the individual level it consists of 2 posttest counseling sessions and 3 skill-building support groups for HIV positive IDUs. The study approach is innovative for 4 reasons: 1) It addresses structural and individual barriers to risk reduction;2) It addresses stigma, which is a primary obstacle to risk reduction for HIV-positive individuals and for which few community level efforts have been undertaken;3) It promotes sustainability by utilizing longstanding and highly effective grassroots organizations in Vietnam;and 4) It enhances various dimensions of social support in the community ("meals on motorcycles" provides instrumental and emotional support), among non-injecting network members ("person important to me" sessions help non-injecting network members support IDUs in their risk reduction goals), and among peers (support group sessions offer emotional and instrumental support). To accomplish this trial, we will: 1) conduct ethnographic research to culturally adapt our intervention;2) select 8 villages and enroll and interview a cohort of 404 HIV-positive IDUs (index) and one of their HIV-negative injecting partners to determine baseline prevalence of sexual and drug risk factors;3) randomize villages to a stigma- reduction or control arm and within each of the 8 villages, further randomize index participants to enhanced posttest counseling and support sessions or an attention-controlled arm;this final randomization will result in 4 study arms and index participants will be assessed at 3, 6, 12, 18 and 24 months for behavioral outcomes while injecting partners will be followed at 24 months to determine HIV incidence;and 4) compare HIV behavioral risks among index participants in the arm that receives both structural and individual level activities to the control arm. We hypothesize that the 2-level intervention will reduce injecting and sexual risk behaviors of HIV-positive IDUs and provide a potential model for reducing HIV-transmission in Vietnam.